The evolution of hāngi methods
Evidence from early Polynesian settler sites in New Zealand such as Wairau Bar and in coastal Otago Peninsula from about 1280 shows a significant number of large cooking pits or umu which were designed to cook Ti-pore or various species of Cordyline. This was a common east Polynesian practice in the Cook Islands and Society Islands. The remains of large umu have also been found in the Kermadec Islands. The .900 long, carrot shaped tap root was cooked in a large stone lined pit for between 1 and 2 days. Various sources say it was baked while some say steamed. The result was a fibrous mass of sweet pulp that had a bitter after taste. Investigation in Otago shows that most pits were used only once or perhaps twice before making a new pit.
Some sources say the longer the cooking, the sweeter the pulp. This may have been due to the evaporation of the water content.
The distinguishing feature of a Ti- pore was its large size compared to a normal cooking earth oven. Prior to colonisation and the introduction of metals and wire, food was laid out on clean sticks, bark, large leaves and other vegetation to minimize direct contact with the hot rocks and reduce burning. Carved bowls and flat rocks were also used for this purpose. Leaves, sticks and vegetation were used to cover the food and to prevent crushing from the weight of the earth on top. Many different hāngi methods are now used. Wire baskets became widely used in the early 19th century with sacking and cloth replacing leaves and bark as the covering of choice. In the early 21st century, gas heated stainless steel “hāngi machines” are sometimes used to replicate the style of cooking without the need for a wood fire, rocks and a pit.